This invention relates to estradiol derivatives, their syntheses and the preparation of their synthetic intermediates or precursors. In particular, this invention relates to the preparation of certain substituted estradiols which have a specific, preferred sterochemistry.
The term "estradiol", as used herein, refers to compounds having the following general structural formula: ##STR1##
Estradiols similar to the compound I are, generically, estra-1,3,5(10)-triene-3,17-diols. In the above drawing the carbons in the steroids structure are numbered according to the generally recognized nomenclature system for steroids, with the hydroxy substituents located at the 3- and 17-positions. It will be understood that the substituent at the 17 position may have either of two orientations, referred to as alpha and beta. Generally, an alpha substituent is one which projects beneath the plane of the drawing shown above and a beta substituent is one which projects above the plane of the steroidal ring drawing. The same would be true for substituents located at C-16.
Substance II shown below is a form of estradiol generally medically recognized to be of particular importance and has a 17-hydroxy-substituent located beta. A commonly used name for this substance is estra-1,3,5(10)-triene-3,17-beta-diol (17-beta-estradiol or beta-estradiol). ##STR2##
Estradiols are naturally occuring substances whose derivatives have been found to have medicinal use. For example, the 3-methyl ether has been used for replacement therapy in estrogen deficiency. Also, certain radioactively labeled estradiols may be used in estrogen receptor assays. In particular, the tritiated, iodine-125(I-125) and bromine-77(Br-77) labeled substances have been tested.
It is known that estrogen receptors, i.e. binding substrates for estradiols, may be found in certain animal tissues. It has also been found that the presence of estrogen receptors may be connected with certain abnormalities in the tissue. Estradiols, if properly labeled, may be utilized to detect the presence of these estrogen receptors in tissue. The medical profession generally theorizes that these estrogen receptor analyses may be conducted in vitro or in vivo. Further, applicants foresee that appropriately labeled steroids may be utilized to deliver a radioactive nucleus or a particular reactive entity to a site of abnormality in tissue, in order to promote a therapeutic effect. For example, a reactive entity which would be selectively carried to a particular type of cells and which would destroy or modify such cells.
In particular, certain 17-beta-estradiols, which have been substituted at the 16-position, are generally thought to have an affinity for estrogen receptors which is both significant and useful in performing estrogen receptor analyses, including assays and imaging. It is foreseen that numerous 16-substituted-17-beta-estradiols may be of importance, particularly those in which the 16-substituent is a halogen and most importantly when the halogen is radioactive. It is foreseen that both 16-alpha- and 16-beta-substituted-estradiols may be of use. However, for any given substituent, the affinity of the 16-alpha-substituted-17-beta-estradiol, for estrogen receptors, is likely to differ from that for the analogous 16-beta-substituted-17-betaestradiol.
Since both the 16-alpha-substituted and 16-beta-substituted-17-beta-estradiols are foreseen to have utility, it is prefered that methods of syntheses of each be developed. It is preferable that each synthetic scheme yield a desired isomer substantially stereospecifically, so that problems of purification and problems from the differences of affinity of the two isomers for binding sites, are avoided. Thus, two general synthetic schemes are needed, one which provides 16-alpha-substituted compound with very little 16-beta-substituted compound being present, and a second general reaction methodology which yields 16-beta-substituted compound with very little 16-alpha-substituted compound being present.
It is readily seen that it would be most desirable to develop a single synthetic precursor or intermediate from which either the 16-alpha-or the 16-beta-substituted compound can be relatively very rapidly and easily formed. That is, given a supply of the synthetic intermediate a synthesis laboratory could easily prepare whichever 16-substituted compound is desired. It is particularly desirable to have alternate synthetic schemes which are relatively easy to conduct and which are both very rapid and very efficient and which produce a relatively high percentage of the desired final product.
With present technology, two basic methods of detection of labeled estradiols are most available. In one, a radioactive substituent is introduced into the molecule and standard methods of radioisotope detection are utilized to determine the presence of the labeled estradiol in the animal tissue, either in vitro or in vivo. In the other, nuclear magnetic resonance (NMR) methods are utilized to detect certain nucleii, and generally non-radioactive labels may be used. At the present time, only radioisotopic techniques are widely available but it is foreseen that other methods may become more available in the future.
If a radioactive isotope is used as the label, then the stereospecificity of the reactions leading to the synthesis of the 16-substituted-17-beta-estradiol may be critical. Also, efficiency of the reaction, in terms of product yield and the length of time it takes to introduce the radioactive isotope into the molecule and then isolate the desired product for diagnostic or therapeutic use, may be very important.
The importance of the stereospecificity is easily understood. Radioactive labels are very expensive and if the reaction is not sufficiently stereospecific large amounts of the label may be lost in undesired products. Also, if the undesired products are to be discarded there may be problems with dangerous residual waste-product radioactivity. Finally, side products might not be easily separable from the desired isomer, and they can interfere with the certainty of assay and imaging data collected when the product is used in medicinal analyses.
If the product yields are not sufficiently high, much of the radioactive isotope may not be incorporated into a useful product, again wasting expensive isotope.
If the reactions involved in the introduction of the radioisotope into the molecule, when coupled with any further reactions or purifications necessary to isolate the desired radioactive products, are not sufficiently rapid, special problems may be encountered. It will be understood that the radioactive label is constantly decaying; and, if the isotope has a sufficiently short half-life, it must be introduced into the molecule rapidly, and the compound must be isolated for biological use relatively rapidly, or the isotope will have passed through sufficient half-lives to produce so little "hot" or radioactive substrate that detection may be difficult.
In some instances the radioisotope used may be contaminated with small amounts of other radioisotopes of the same compound. For example, Iodine-123 (I-123) as it is currently made, is often contaminated with some Iodine-124 (I-124). The half-life for I-123 is about 13.3 hours whereas the half-life for I-124 is about 4.2 days. I-123 is readily detectable and gives clear images whereas I-124 generally causes some scattering and images of low resolution. Consider what happens if a mixture of 99 to 1, I-123 to I-124, is utilized to label a substrate. If the reaction takes too long, for example 26 hours, for introduction of the isotope into the substrate molecule and isolation of the desired product, then the I-123 will pass through about two half-lives and only about 25% of it will be left. The I-124, however, will have barely begun decaying and nearly 100% of it will be left. The ratio of I-123 to I-124, after the 26 hours, will have changed to approximately 24 to 1. It is readily seen that this enhancement of the amount of I-124 present, by ratio, may cause difficulty since the I-124 might make resolution of images difficult. Also, if much of the isotope mixture must be given to accommodate imaging of I-123, residual radioactivity from the I-124 component, with its long half-life, may be a problem. These types of problem are usually present whenever an isotope of short half-life is used, if the isotope is normally contaminated by a second isotope of longer half-life.
In some instances, the decayed product may still be active as far as an estrogen receptor is concerned and the labeled, but no longer hot, estradiol derivative may block estrogen receptors from receiving the hot substrate, thus interfering with the accuracy of any assay or imaging data obtained. This problem cannot be overcome by using an excess of the estradiol material since doing so may tend to overload the estrogen receptors and send hot estradiol to other locations, where it may be detected, generating erroneous conclusions about the presence of estrogen receptors.
When non-radioactive isotopes are used in chemical syntheses, problems of low yield and low specificity are often overcome by utilizing large amounts of starting materials, and labels, and undergoing sufficient purifications to allow for the isolation of significant amounts of the desired products. It is clear that this methodology is generally unacceptable when radioactive labels are used. First, radioactive labels are usually too expensive for an inefficient synthesis scheme to be commercially utilizable. Secondly, radioactive isotopes can be dangerous and large concentrations of them should be avoided. Also, unreacted starting materials and undesired side products may be radioactive, causing problems with contamination during clean-up, isolation and waste material disposal. Further, if the isolation of the desired product takes too long there may be problems with decay of the isotope.
Radioactive isotopes of the halogens are generally considered to be the most important types of labels for use in labeling compounds for biological assays and imaging. The isotopes generally considered to be of most importance are fluorine-18 (F-18), bromine-77 (Br-77), iodine-123 (I-123) and iodine-125 (I-125). Fluorine-18 is a positron emitter and has a half-life of approximately 110 minutes with a relatively high energy of decay. Detectors for positron emission are not widely available at the present time, so, while fluorine-18 may in the future become widely used as an isotope label, its potential has yet to be completely realized. It is foreseen, however, that fluorine-18 compounds may be made according to the present invention, so as to provide the compound when the demand therefor increases.
Bromine-77 decays with gamma emission and has a half-life of approximately 56 hours.
Iodine-123 is a gamma emitter having a half-life of approximately 13.3 hours. Its energy of gamma decay is relatively high, approximately 159 kilo-electron-volts (KeV). The toxicity of iodine is generally well understood, and I-123 is generally considered to be an almost ideal radioisotope for use in biological studies. In particular, its relatively short half-life makes radioactivity contamination a relatively minor problem, while at the same time, its relatively high energy of decay makes detection relatively easy, even in vivo.
Iodine-125 has a half-life of approximately 60 days and is a gamma emitter having a gamma decay energy of approximately 35.48 KeV. Its relatively long half-life makes it undesirable for many uses since there will be residual radioactivity for a considerable period of time. Its relatively low gamma emission energy makes it hard to detect, with most instruments, especially when used in vivo, although even in vitro detection can be a problem.
While there are many other radioisotopes of the halogens, the above discussed radioisotopes are the ones most generally considered for use in diagnostic medicine. Of them, I-123 is presently considered to be the most generally desirable and practical isotope for use in labeling estradiols.
In the past, 17-beta-estradiols labeled with I-123 at the 16-position have been unavailable in amounts and purities generally considered to be useful in assays, and other diagnostic work, either in vitro or in vivo, due to problems in their syntheses. Generally, these problems result from the length of time formerly required to introduce an I-123 label into the 16-position, stereospecifically, and the length of time required to complete the synthesis and isolate and purify the desired product. In addition, those synthetic methodologies which were considered in the past were often of low yield and often resulted in the waste of large amounts of I-123 label.
The advantages of the present invention, in preparing labeled compounds, will be most apparent if an examination is first made of previously known methods of introducing halo-substituents into the 16-position of 17-beta-estradiol. Such an examination, of the major known methods, follows:
A highly publicized method of preparing 16-alpha-halo-substituted-17-beta-estradiol is that published by R. B. Hochberg and will be generally referred to as Hochberg's method or synthesis. The final step of Hochberg's synthesis is shown below and comprises Iodo-substitution on the 16-beta-bromo-compound, i.e. a Finkelstein reaction: ##STR3##
In this spcification and the claims: a wedge indicates a substituent projecting above the plane of the drawing; a dotted line indicates projection below; and, a curve indicates a mixture of both. These are conventional methods of indicating stereochemistry. Also, Ac is used to indicate an acetyl group, --C(O)CH.sub.3 .
The substitution reaction is generally run in 2-butanone for anywhere from 12 to 24 hours. Although the reaction has been utilized to introduce the radioisotope I-125 into the 16-alpha position of 17-beta-estradiol, it is generally considered to be of too low a yield and too long a length of time to allow efficient production of a compound by the substitution of a radioisotope, such as I-123, which has a relatively short half-life.
Even if conditions are found which allow for an increase in the rate of the substitution reaction, it appears unlikely that the synthesis and purification method proposed by Hochberg can be readily and economically utilized to prepare commercially useful 16-alpha-substituted radioactive estradiol derivatives when the radioactive isotope has a very short half-life. For example, the Hochberg synthesis may require time-consuming product isolations and purifications. Further, the reaction does not appear to be adaptable to substantially stereospecific preparation of 16-bata-substituted compound.
The following scheme shows the overall Hochberg method of synthesis: ##STR4##
There are certain problems with the above reaction scheme. For example, the bromination step yields two compounds, mostly the beta form, and the reduction step gives a mixture of all four possible bromohydrins, from which the desired isomer has to be isolated. No single intermediate is formed from which the 16-alpha-substituted compound can be rapidly formed and from which the 16-beta-substituted compound can also be rapidly formed. Also, the starting material for the final Finkelstein, 16-beta-Br-17-beta-estradiol, and the product of the final substitution, 16-alpha-I-17-beta-estradiol, have similar characteristics for chromatographic purposes, so their clean separation from one another, in the event that the final substitution does not go to completion, can be difficult. Further, during the Finkelstein reaction and work-up there may be epimerization of starting material or product, thus decreasing efficiency. Also, one epimerization product, 16-alpha-Br-17-beta-estradiol, has very similar chromatographic properties to the desired product, making purification somewhat difficult.
Accordingly, researchers' initial attempts to form 16-alpha-.sup.125 I-17-beta-estradiol via the Hochberg method resulted in products having specific acivities which were very low, approximately 95 to 140 Curies per millimole (Ci/mmole), as opposed to the theoretical specific activity of approximately 2,000 Ci/mmole. It is reported in the literature that meticulous purification of the 16-beta-Br-17-beta-estradiol used in the Finkelstein reaction has resulted in some increased specific activity; however, purity of products still appears to be a problem and seems to keep specific activity down.
Another, practical, problem is associated with Hochberg's method. Most radioactive halogen anions are commercially available in the form of an ammonium salt or a sodium salt. In the case of radioactive iodides both ammonium salts and sodium salts are usually available whereas, in the case of the bromides and fluorides, usually only the sodium salt is available. These radioisotopes are normally shipped in water, which does not readily evaporate, and significant amounts of base, either sodium hydroxide or ammonium hydroxide, will be present. The Finkelstein reaction, and its starting material, may be sensitive to the presence of either base or water. Therefore, the commercially available isotope, in its basic storage mixture, usually must be scrupulously neutralized and dried before it can be utilized. It is evident that it would be desirable to develop a reaction methadology which is at least relatively insensitive to the presence of water and preferably can tolerate base.
An alternative method of labeling 17-beta-estradiol with halogens at the 16-position has been developed and utilized by Katzenellenbogen et al. Their general reaction methodology is shown below: ##STR5##
Katzenellenbogen et al. J. Med. Chem., 23, p. 994-1002(1980) describe bromination of the enol acetate (1) followed by hydrolysis to give 16-alpha-Br-3-hydroxy-1,3,5(10)-triene-17-one(16-alpha-Br-estrone; (2b) as proceding with relatively high yield and relatively high specificity for the alpha-product (2b). Formation of the 16-beta-Br-compound, 4a, was accomplished by epimerization of the 16-alpha-compound (2b) in acid. The epimeric ratio was 1/1.8 (2b/4a). In commercial use, such a mixture would have to be separated by chromatography and clean separation can be expected to be difficult and time-consuming.
16-alpha-Br-17-beta-estradiol (3a) was formed from reduction of the ketone (2a) with lithium aluminum hydride. (LiAlH.sub.4). The reduction yields both the 17-alpha-and 17-beta-alcohols (3a and 3b) in a ratio of 2/1 (17-beta/17-alpha; 3a/3b). Such a mixture can, in theory, be separated by chromatography; however, again, since the products are so similar, chromatographic separation may be expected to be difficult and time-consuming.
Katzenellenbogen prepares Hochberg's precursor, compound 5 above, by reduction of a mixture of 16-alpha-and 16-beta-Br-estrones with zinc borohydride (ZnBH.sub.4). All four bromohydrins are formed from such a reduction, since the reduction is not stereospecific. Again, although chromatographic means may in theory be utilzed to separate the four bromohydrins, they are similar enough in properties so that the separation can be expected to be very difficult.
16-alpha-Cl-estrone (2c) was formed by Katenellenbogen et al., by treating the enol acetate (1) with tert-butyl hypochlorite. A mixture of products is generally formed from such a reaction so some purification is necessary for isolation of the chloride.
16-alpha-Cl-17-beta-estradiol (3c) and 16-alpha-Cl-17-alpha-estradiol (3d) are formed from treatment of the estrone (2c) with LiAlH.sub.4. Isolation of either product requires purification generally by chromatography, which can be difficult with such similar reaction products.
It may be appreciated from the above that the variation of the Hochberg method utilized by Katzenellenbogen et al. is generally inefficient for 16-substituted compounds, especially when radioisotopes are being used. First, in nearly all instances, product mixtures are formed, which wastes expensive label and which can be hard to separate. Also, too much time may be needed to easily handle radioisotopes of relatively low half-lives.
Longcope et al. have developed a method of synthesis for 16-beta-I-17-beta-estradiol which is significantly different from Hochberg's synthesis and the Katzenellenbogen variations. Their synthesis begins with 16-alpha-17-beta-estriol and is shown below: ##STR6##
The Longcope method yields a product mixture which includes the desired beta-product together with an elimination product, so, again, it is inefficient. Secondly, it cannot be readily adapted for formation of 16-alpha products. Also, radioactive triphenylphosphite methiodide is expected to be difficult to prepare. In addition, it does not appear to be easily adaptable to the utilization of other halogens and their radioactive isotopes.
The above three general synthetic methodologies illustrate many of the problems associated with syntheses of 16-substituted, either alpha- or beta-, 17-beta-estradiols. Generally, reaction mixtures including numerous products are formed. Also, no synthesis is readily adaptable to yield, stereospecifically, either the 16-alpha or the 16-beta isomer as desired. Also, the lengths of time required for the introduction of halogen label and isolation of the products tend to make the utilization of the above schemes for radioactive isotopes, especially isotopes with relatively short half-lives, very difficult, if not impossible.